Nature of Science

Science is a dynamic, evolving process. What we consider 'scientific knowledge' today may be revised tomorrow as new evidence emerges or better technology becomes available.

A key example is the shift from Spontaneous Generation (the idea that life arises from non-living matter, e.g. maggots from rotting meat) to Germ Theory (diseases are caused by specific microorganisms). Louis Pasteur's famous swan-neck flask experiment in the 1860s finally disproved spontaneous generation.

Similarly, the Miasma Theory ('bad air' caused disease) was replaced when scientists like John Snow traced a cholera outbreak to a contaminated water pump in London (1854), showing disease spread through water, not air.

Accidental discoveries (serendipity) have transformed science - Alexander Fleming noticed mould killing bacteria on a petri dish in 1928, leading to the discovery of penicillin, the first antibiotic.

Technology drives science forward: the microscope revealed cells (Robert Hooke, 1665), and modern DNA sequencing has unlocked genomics and personalised medicine.

Science is not done in isolation - it is a global, collaborative enterprise. Researchers worldwide share findings through journals, conferences, and digital platforms.

The Human Genome Project (1990–2003) is a landmark example: 2,800+ scientists from 20 institutions across 6 countries worked together to sequence the entire human genome (3 billion base pairs). This collaboration revealed that humans share 99.9% of their DNA.

Modern science increasingly relies on international collaboration - COVID-19 vaccine development involved scientists from dozens of countries sharing data in real time, leading to vaccines in under 12 months (compared to the usual 10-15 years).

Before any research is published in a scientific journal, it undergoes peer review - a rigorous evaluation by independent experts in the field.

The Peer Review Process:

1. Researcher submits paper to a journal
2. Editor sends it to 2-3 independent experts (reviewers)
3. Reviewers check: experimental design, data analysis, conclusions, bias, plagiarism
4. Reviewers recommend: accept, revise, or reject
5. If accepted after revisions, the paper is published

Repeatability vs Reproducibility:

• Repeatability: the same researcher, under identical conditions, gets the same results
• Reproducibility: different researchers, in different labs, following the published method, get the same results

A test-retest correlation of +0.80 or above indicates good reliability. To improve reliability: publish detailed methods, use repeated trials, and increase sample size.

Bias is any systematic error that skews results away from the truth. It can occur at every stage of an investigation - from sample selection to data analysis to reporting.

Types of Bias:

• Sample bias: the sample doesn't represent the population. E.g. surveying only one school about favourite subjects
• Confirmation bias: researchers only look for evidence that supports their hypothesis
• Analysis bias: fabricating data, omitting inconvenient results, using misleading graphs
• Publication bias: journals prefer to publish positive results, not negative ones

Recognising Graph Manipulation: A truncated y-axis (one that doesn't start at zero) can make small differences appear dramatic. Always check the axis scale when reading scientific graphs.

Reducing Bias: Use random sampling, blind/double-blind designs, peer review, and check: Who funded the study? Are all results reported? Is the source peer-reviewed?

The scientific method is the systematic approach used by scientists to investigate questions about the natural world. It provides a logical framework for testing ideas and drawing conclusions.

A hypothesis must be:
• Testable - can be investigated through experiment
• Falsifiable - can potentially be proven wrong
• Written as a clear, predictive statement (not just a guess)

Variables explained:
• Independent variable (IV): what the scientist deliberately changes
• Dependent variable (DV): what is measured in response
• Controlled variables (CVs): everything kept constant for a fair test

Example: Testing if caffeine affects heart rate - IV: dose of caffeine, DV: heart rate, CVs: age, gender, resting period, time of day.

Controls in Medical Trials:
• Control group: receives no treatment (or a placebo)
• Placebo: a fake treatment that looks identical to the real one
• Double-blind design: neither participants nor researchers know who receives the real treatment - eliminates expectation bias from both sides

Understanding measurement quality is critical. Three related but different concepts:

Accuracy = how close your result is to the true value. Like hitting the bullseye on a target.

Precision = how close your repeated measurements are to each other. Like hitting the same spot repeatedly (even if it's not the bullseye).

Validity = does the investigation actually measure what it claims to? A valid experiment has proper controls, a representative sample, and calibrated instruments.

Think of it as a target:
• High accuracy + high precision = all arrows in the bullseye
• High precision + low accuracy = arrows clustered together but off-centre
• Low precision + low accuracy = arrows scattered everywhere

Data can be qualitative (descriptive observations like colour change, smell) or quantitative (numerical measurements like temperature in °C, mass in grams).

Anomalous data (outliers) are results that don't fit the pattern. They should be excluded from averages but always reported and explained - never deleted silently.

Two types of error:

• Systematic error: a consistent, predictable deviation from the true value. Affects accuracy. Caused by faulty equipment, calibration errors, or flawed method. E.g. a balance that always reads 0.5g too high.
• Random (statistical) error: unpredictable variation between measurements. Affects precision. Caused by natural variation, human reaction time. Reduced by increasing sample size and repeating trials.

Evaluating results is a core scientific skill. After collecting data you should:

• Compare the data to your hypothesis - was it supported, partially supported, or refuted?
• Identify sources of systematic and random error and estimate their likely impact.
• Suggest specific improvements - e.g. more precise instruments, a larger sample, better controls.
• Compare your results to published literature values or pooled class data to assess consistency.

Communicating results takes many forms, each with a different purpose and audience:

• Peer-reviewed journals - detailed articles describing methods, results and conclusions; evaluated by expert reviewers before publication.
• Conferences - presentations and posters that enable face-to-face discussion among scientists.
• Preprints - papers shared online before peer review to speed up communication (common during COVID-19).
• Popular media - news articles, blogs and social media that translate findings for the public but may oversimplify.

Always work safely and ethically: complete a risk assessment, wear appropriate PPE, dispose of waste correctly, and cite the work of others honestly.

Work through each question in full sentences, then check your answer against the model answer and marking points that follow it.

Question 1 (15 marks) - A student investigates whether a new fertiliser increases crop yield.

1. Write a suitable hypothesis for this investigation. (3)
2. Identify the independent and dependent variables. (4)
3. Name two controlled variables and explain why controlling them is important. (5)
4. The student repeats each condition three times and calculates an average. Explain how this improves the reliability of the results. (3)

Question 2 (14 marks) - A researcher publishes a study claiming that a new probiotic supplement reduces blood pressure in adults aged 40-50.

1. What is meant by 'peer review'? State one benefit of the peer review process. (4)
2. The study used a sample of 12 participants, all from the same city. Suggest two reasons why these results may not be reliable. (4)
3. Distinguish between a systematic error and a statistical (random) error. Give one way to reduce each type. (6)

Question 3 (11 marks) - A study finds a positive correlation between the number of hours students spend on social media and their reported anxiety levels.

1. Explain the difference between correlation and causation. (4)
2. Suggest one confounding variable that might explain this correlation. (3)
3. Describe how a controlled experiment could help establish whether social media causes anxiety. (4)